sage.gms
  main sage model



--------------------------------------------------------------------------------
running the model
--------------------------------------------------------------------------------

the model is designed to run a single scenario, whether that is the baseline or
a policy case. this is intentional, first to reduce the runtime by not requiring
the model to solve for the baseline each time it is run, second to reduce the
size of the output files by nearly a half.

the model is intended to be run from the model's top directory.

running the default baseline simulation is done with a simple call to GAMS:

  gams model/sage.gms

to define the output file use the --output_file flag. for example to write the
output to the file output_file.csv in the output directory run:

  gams model/sage.gms --output_file=output/output_file.csv

to run a policy scenario the user must supply a GAMS file that implements the
policy using the --policy_file flag. for example to use the example_shock.gms
policy in the experiments subdirectory run:

  gams model/sage.gms --policy_file=experiments/example_shock.gms

to run sage using different parameters use the --paramter_file flag. for example
to run the static version of the model's baseline run:

  gams model/sage.gms --parameter_file=data/parameters_static.gms

by default sage saves the model solve in a gdx file output/results.gdx. to
change this location use the flag --gdx_results_file. for example to save a
baseline file in the output directory under the name baseline.gdx run:

  gams model/sage.gms --gdx_results_file=output/baseline.gdx

a gdx file of a model solve may be used as a baseline for a model run to both
set the starting values and to define baseline welfare and prices to calculate
welfare impacts using the --gdx_baseline_file flag. for example, to use a
baseline solve stored as output/baseline.gdx run:

  gams model/sage.gms --gdx_baseline_file=output/baseline.gdx

for an example of how to use the model and its results from R see the file
basic_example.R in the experiments subdirectory
